This final chapter is concerned not so much with the nature of the Celtic gods themselves as with some of the distinctive methods by which Celtic artists symbolised and portrayed cult-material. The character of Celtic sacral art was touched upon in Chapter One and we here examine the specific traditions which develop this theme. Imagery gives clues as to how the gods were envisaged and worshipped. In addition, an examination of the possible purposes underlying the use of varying art-styles reveals a level of sophistication on the part of both craftsman and patron which belies the common definition of much of Gaulish and British indigenous cult-portrayal as ‘primitive’ and ‘incompetent’.
Celtic sacral depiction, whether of human, animal or inanimate subjects, frequently exhibits features which transform ‘real’ imagery (that is copied from life) into something deliberately less life-like. This was achieved in a number of ways, but all have the dual effect of removing the image from the world of reality and – in some instances at least – of augmenting its power as a symbol. This Celtic iconographic tradition (which must of necessity be studied mainly from Romano-Celtic data) may display a high degree of stylistic abstraction or schematism. Very often an abbreviated ‘shorthand’ method of representation is employed not only in stonework (though such is the commonest medium) but also occasionally in metal. The bulk of such iconography is religious in theme. It is here argued that the summary dismissal of these depictions, by ‘modern’ criteria, as being of inferior workmanship is quite inapplicable: they must rather be examined on their own terms of function and context. Schematised representation was, indeed, a conscious and successful form of image-making. The Graeco-Roman ideal of mimesis was frequently rejected, not because Celtic artists were unable to produce faithful copies but because rigidity was, on the one hand, unnecessary within the context of cult imagery and, on the other, was perhaps viewed as inappropriate for the divine. The same tradition which produced schematised images of deities in human or animal form, saw no reason not to break the rigid Graeco-Roman framework of realism in other ways. Of these the most important appear to be emphasis achieved either by multiplication/addition or by exaggeration of one element or, at its most extreme, by the representation of pars pro toto. Enhancement by addition may either be a straightforward multiplication of the whole or part of the image, or it may involve the introduction of a new, alien element or attribute to form a composite depiction. Triplication of Celtic mother-goddesses is an example of the former and ram-horns added to a depiction of a serpent of the latter. As well as by plurality, emphasis may be created by the exaggeration of an otherwise normal part of an image. This is exemplified most clearly in cases where the head has been enlarged out of proportion (culminating in the representation of the head alone). Conversely, other less important parts of the body may be shrunken to create more (or less) emphasis.
It is apparent that there are elements common to the different aspects of emphasis outlined above. Most important is the conversion of a ‘normal’ human or animal image to something which is overtly supernatural on account of its deliberate depature from the realistic, a practice which had nothing to do with style or artistic expertise but rather with deliberate choice of content. It is advantageous here to stand back from the evidence and look conceptually at the phenomenon of plurality and exaggeration. Both forms of manifestation may be treated together since both appear to serve the same purpose of enhancing the potency of the symbol and of ‘hammering home the point’ both perhaps to the divinity honoured and to the worshipping devotee. In this chapter we will be examining the reasons for overemphasis and, in that context, we shall raise the question of the role of symbolism in Celtic iconography, an issue equally relevant to the phenomenon of schematism. The final type of portrayal we will look at is a related phenomenon, that of miniaturisation. Though the manufacture of model objects was not Celtic in origin, the practice seems nevertheless to have been adopted par excellence for Celtic religious purposes. It will be argued that – notwithstanding arguments concerning convenience and cost – there is a specific symbolism associated with the offering of perfect miniature replicas of everyday tools and weapons to supernatural powers.
The issue upon which success or failure of image-making must be judged consists both of the sculptor’s or craftsman’s intention and of the function of the created image. The production of a religious representation may serve several different purposes, not necessarily all mutually exclusive. If a stone carving which portrays a deity is placed in a temple, shrine or locus consecratus, it may possess a function separate from one which comes from a domestic context or a grave. A temple-depiction may be the cult-image: this means that, originally at least, the god would be considered as dwelling within that image. Alternatively, the carving may be seen not as housing the god specifically but may have been set up in order to demonstrate reverence and honour to the god by symbolic physical representation in much the same way as the erection of a statue may commemorate a revered human individual. More prosaically, an image may be created and established in a shrine as a focus of worship, to channel the attention of the devotee and to stimulate thoughts about the divine. All these functions may be true especially of the larger, more monumental, representations. If, however, we turn to small objects (usually of metal), two categories may be recognised: first liturgical material associated with cult-practice or ritual and, second, personal items. The first group occurs generally in temples or buried in caches (presumably originally in shrines or sacred places) and may be regarded as the sacral property of priests or religious officials. The second group may be viewed as generally belonging to individuals, and examples of these occur in shrines as offerings, in houses or in sepulchral contexts. The function of ritual material is more or less self-evident: it may be assumed that a ceremonial sceptre or priest’s mask and headdress bearing divine images is itself imbued with sanctity or that holy power is invested in the object by means of the divine portrayal. The interpretation of personal items is more complicated. Symbolism of the divine must in all instances be present to a greater or lesser extent; small bronze or clay figurines must have been purchased and dedicated to the deity represented. But whilst some would have been dedicated ‘officially’ and placed in roadside, urban or domestic shrines, others, such as amulets, might remain in secular contexts and be regarded rather more vaguely, as good-luck symbols or talismans rather than as possessing any genuine sanctity as a result of deliberate consecration.
Finally, in arguments concerning the function of symbolism or a symbolic object, it is necessary to clarify the responsibilities involved in the production of cult-images as visual expressions of belief. In each instance, two distinct individuals were generally involved (perhaps more if epigraphic dedications are present), namely the initiator and the executor – the patron and potential purchaser and dedicant on the one hand, and the craftsman on the other. I would argue that both are key figures in determining the type of image produced, for both patron and stonemason/ craftsman need to be in tune with the form of image required. A patron would be unlikely to commission/ purchase a cult-image whose iconography he did not understand. Likewise, a craftsman would need to be au fait with the style required of him. This means that symbolic content would have been a careful matter of conscious choice by people who were fully aware of what means would best be employed in endowing a cult-object with the greatest possible potency.
Study of stone (and to a lesser extent) of bronze iconography of any Romano-Celtic province reveals a striking diversity of stylistic treatment in the imagery expressed. Whilst many religious depictions in the pagan Celtic world made during the Roman occupation display classical realism, this is balanced by a significant proportion of cult-images which owe little to Graeco-Roman artistic traditions of verism or naturalism. Some portrayals indeed, exemplified by the chalk figurine from Deal, Kent or the Ivy Chimneys carving, give the impression of minimum attention to the accuracy of the human or animal form, often being little more than a sketched outline, apparently of much the same standard of artistic expertise – to a modern art-historian’s eye at least – as a child’s drawing.
Stylisation, schematism or abstraction appear to be Celtic traits. However, whilst there may be some correlation between the ethnic origins of the deities represented and the art-form in which they are depicted, this is by no means the norm. The Emberton Mercury-relief is identifiable as Mercury because of the attributes of wings and caduceus but the style of representation is entirely Celtic. Conversely, the depiction of a specifically Celtic divine concept such as the Matres from Cirencester may on occasions possess a high degree of realism, indicating a faithful attempt at naturalistic depiction of the female human body. Detailed examination of a cross-section of Romano-British iconography appears to reveal (in terms of art-treatment) not two but three main categories of image. The naturalistic and the highly schematised iconography represent opposite ends of the spectrum of style. But there may be distinguished also a medial category which poses greater interpretive problems. This group consists of iconography displaying apparent (the italics are crucial) attempts at, but incompletely successful production of, the realistic human or animal figure. This category is exemplified by the badly-proportioned Mars-Romulus relief from Custom Scrubs and the horseman from Stragglethorpe, Lincs.
Schematism must be assessed within the context not of art for art’s sake but of religion. To interpret expertise or the lack of it merely from a present-day art-analyst’s point of view is to argue from an entirely false premise, for the Celts’ lack of realism may be deliberate, with obscurity and the use of shorthand depiction serving almost as a divine ‘code’ with the specific purpose of relating the image directly to the supernatural. At the very least, even if not always deliberate, the medial category of semiveristic portrayals may represent the lack of necessity to treat mimetically the human image of a deity. In order to understand this, it is useful to look at the principles of classical and Celtic art.
A fundamental factor in assessing the excellence or otherwise, in artistic terms, of schematised Celtic religious representations, is the recognition and acknowledgement of the existence of specific ethnic artistic traditions. Critics of ‘crude’ Celtic iconography appear to judge entirely from the classical standpoint that realism or mimesis equal excellence. This can be appreciated by examining some of the basic differences between the standards and traditions of Celtic and classical representative art.
In the classical world, the artist’s main function was to please or impress both gods and men; his was an imitative rather than an interpretative role. The aim was to portray man at his idealised best, indistinguishable from the gods represented by him in man’s likeness. The naturalism of Greek art in the Classical period was motivated and stimulated largely by the wish to depict human beings (whether gods or men) as real, close-to-life and thus more comprehensible to the beholder. The Greeks were obsessed with the true nature of appearance; indeed the imitation of nature, ‘mimesis’, was introduced by the Greeks, who escaped from ‘conceptual’ images and became true to life. In the heyday of Classical Greece art was a curious and uneasy blend of idealism and realism ‘. . . idealised mortal is near-divine, self-sufficient and above ordinary passions’ (Boardman 1973, 120). But by the fourth century BC a new trend can be observed: features became softer and more relaxed and emotion, where appropriate, crept in. If we look at Roman Republican portraiture, we are confronted with stark realism or verism, derived in part from Etruscan influence. The Roman attitude to artistic excellence is interesting in that the degree of realism achieved is termed successus. Pollitt points out that Vitruvius disapproved of the fantastic wall-paintings of his day because they represented things which could not exist in nature.
Celtic art sprang from totally different roots and intentions from those of the Graeco-Roman world. Unless the schematic cult-images of the Celto-Roman world are assessed within this different tradition, criticisms may be made which are both irrelevant and inappropriate. I would argue that, in the light of observations made below, such comments as ‘. . . one wonders whether all are by the same incompetent hand’ (Rhodes 1964, 32ff) or ‘some of the carvings from the Cotswolds . . . are very simple’ (Henig 1984, 62) are perhaps themselves both simplistic and meaningless.
Celtic figural representation reflects general Celtic artistic tradition which, in the main, was an abstract, symbolic tradition. The art of ‘barbarian’ Europe was ‘symbolic and enumerative, presenting a series of representational clues which can be mentally assembled’. It is interesting that critics of Celtic iconography within the Roman period speak with scorn of the crude images present and at the same time postulate the death of true Celtic art under Roman influence. It may rather be that Celtic tradition is quite logically manifesting itself traditionally but in response to the stimulus of the Roman emphasis on the human or animal image of the supernatural. Prehistoric Celtic art (which predates the period of classical influence) was essentially decorative; patterns and the embellishment of surfaces were of prime importance.
SCHEMATISM: FUNCTION AND REASON
In examining the stimulus behind abstraction in Celtic humano-divine representation, it has been argued that insular Celtic art differed markedly from the Graeco-Roman tradition in not being man-dominated. Certainly there was no need for a predominantly rural and farming-oriented society to be preoccupied with the realistic. Even where human depiction was present, there may well have been deliberate attempts by craftsmen at de-humanising. Celtic sculptors (and presumably their patrons) did not identify their gods in terms of human perfection, which would very possibly have been largely irrelevant to Celtic notions of divinity, rooted in natural phenomena and sensed rather than envisaged and directly perceived. It is possible to argue deliberate choice on the part of Celtic artists who were producing stylised figures, not lack of ability and artistic skills. Anati believes that the most schematic of the Val Camonica rock-carvings (which date from the Neolithic to the Iron Age) are those with a religious theme. The avoidance of the realistic human element did not mean that Celtic art was necessarily crude (in the pejorative sense), primitive or inferior, but merely different, stimulated by principles other than mimesis. This may be paralleled in other prehistoric symbolism. In the Neolithic Vinc˘a culture of Yugoslavia for instance, schematism in art was demonstrably deliberate: the artist produced something aimed not at aesthetic effects but purely at the symbolic and conceptual. This prehistoric schematism or formalised reduction should not be ascribed to technical ineptitude but to skilled craftsmanship conforming to matured traditions, concepts and beliefs. In old Europe in the seventh and sixth millennia BC abstract and naturalistic forms appear side by side at one and the same time.
Celtic craftsmen may have used schemata (i.e. limited or limiting forms) where geometry rather than nature was the inspiration, it may be that Celtic art was especially susceptible to schemata or ‘minimum clues of expression’. If one abandons the Graeco-Roman criterion for artistic success – realism – the definition of an artist is not one who merely copies lifelike forms as slavishly as possible, but one who acknowledges a theme by means of a physical expression which cannot be achieved by realism. Representation in art does not have to be naturalistic in order to be good; artistic value depends on the presence of a formal element not identical to the form found in nature. One would not equate photography with art simply because it copies faithfully, and that is the highest realism possible. Indeed, the Celtic artist displays a sophisticated genius in his ability to look at a model – perhaps a human figure – and reduce the representational essentials to the stark and manageable few, enough for recognition but no more. Thus, economy of detail captured the essence of a figure or face with the minimum of physical expression. For an image to be successful, it may need to be no more than a mere scratching so long as it retains the ‘efficacious nature of the prototype’. The Ivy Chimneys figure is very stylised and deceptively simple, but it is evocative and was carved with some care.
In judging the success or failure of Celtic figurative art, one returns again and again to function and to the realisation that such art is predominantly religious. The form of representation must be assessed in terms of the purpose and requirements of a particular society. In a sacral context, the choice of a particular style of representation may have significance. Allusion has already been made to the non-recognition of a human ideal in deity-depiction and, where gods are portrayed in human form, the stick-like, shorthand, ‘careless’ method of depiction – seen for instance on the Farley Heath sceptre-binding – may be deliberate, due partly to the wider field of general Celtic art-tradition, but also to conscious obscurity, enigmatism and ambiguity. Religious representations may not necessarily be intended as works of art and should not be so judged. Form and exactness may not have been relevant. There may have been no belief that a schematised figure was an accurate reflection of the object represented. Abstraction and schematism can arguably play as important a role in art as naturalism. The geometric pattern of the Genii cucullati from Cirencester must surely be inspired by something other than a feeble attempt at three cloaked and hooded figures. Here the positive and negative elements of the carving may both be important: the figures themselves and the spaces between them seem carefully and sensitively balanced, as if both abstract pattern and the human image are inextricably blended in terms of intended symbolism, and the essential detail of threeness and hoodedness are very apparent. Veiled, obscure interpretation and ambiguity will be further examined below. I turn now to emphasis – the other, related, theme of this chapter.
EMPHASIS: TRIPLISM
Multiplication of the cult-image is perhaps the commonest form of visual emphasis and in Britain, as in Gaul, the most usual form taken is that of triplication. The representation of deities in threes is not specific to any one god-form, but occurs some of the time in depicting particular divinities. Certain forms – such as the Tres Matres are defined by their triple form; but single mother-goddesses occur also. Genii cucullati are traditionally represented in triplicate in Britain, although in German contexts they appear not as three dwarves but as single giants. Other British examples of triplication include the triple Mars from Lower Slaughter and the triple genius from Symonds Hall Farm (Glos.). In Gaul a triple-faced form was common among the Remi of the north-east where eleven such images are recorded and where even coinage depicts this form. The occasional British three-headed image is exemplified by a carving from Wroxeter, the three-faced head from Bradenstoke, Hants. and the triple head from Corleck, Cavan. Rather different manifestations of the same basic phenomenon include triplication of horns on bull-figurines, common in eastern Gaul (Chapter Six); the triple phallus on a Gaulish Mercury and the three horns on a boar-figurine are further examples of the ubiquity of triplication in various forms. In the vast majority of instances where triplism is present, it is evident that the number ‘3’ is important and so is likely to have possessed powerful symbolism for the Celtic people. The archaeological evidence is supported by early Welsh and Irish literary tradition where triads are prominent. But it is demonstrable that triplism is not the whole story; multiplication is significant in itself. This is displayed both by the occurrence of double or Janus-heads, like those from London and Iron Age Holzerlingen, and by the occasional instance of quadruplism. This last evidence is tenuous: it is arguable that the Wroxeter carving may represent either three or four heads, and the quadruple mother-goddess carving from London may, in fact, represent the Tres Matres and the Empress Julia Domna (either the living woman herself or the deified version of the deceased empress).
In examining the practice of plurality in iconography, we may identify two related elements – the use of the number ‘3’ and the presence of more than one image. The two ideas are not mutually exclusive but do exhibit certain differences of outlook or stimulus. We have to consider whether ‘3’ possessed power over and above the conceptualisation of a deity with three facets to his/her cult. The Tres Matres, for instance, could be interpreted as symbolic of life, death and rebirth, associated with the seasonal cycle of the earth’s fertility. Alternatively, the triple aspect may be linked with childhood, adulthood and old age; or the fecundity of people, beasts and crops – we can only speculate as to the possible different symbolism of each of the three images. There may be more to it than that: whilst some representations of the Matres are slightly different one from the other, as at Ashcroft, Cirencester, and in the Rhineland, or may bear such different attributes as children, bread, fruit or dogs, other triple depictions, such as the Lower Slaughter triple Mars, are identical with each other. Most important, however, is that, of all multiplication, by far the commonest factor is ‘3’.
EMPHASIS: EXAGGERATION BY ENLARGEMENT
Exaggeration in symbolism is found as far back as the Upper Palaeolithic, where we may cite the Gravettian ‘Venus’ figures; here facial details are disregarded and the concentration is entirely on the imagery of fecundity and prosperity. Exaggeration is frequent too in the Bronze Age rock-art of Scandinavia, southern France and North Italy, where the image is schematised but where important attributes such as horns on cattle, weapons and the hands of men are over-emphasised.
In the Romano-Celtic world this most frequently manifests itself in the over-emphasis of the human head, as in the squatting wheel-god from Churcham, Glos., the Epona from Albaina in Iberia or the cross-legged bronze figure from Bouray in northern Gaul. In beasts, the stress may instead be on the horns of bulls and goats, the antlers of stags or, in the case of boar-figurines, the dorsal bristles may be depicted larger than normal. Exaggerated horns occur on a number of Iron Age and Roman bronze bull-head bucket-escutcheons and on a recently discovered bronze goat from south-west Scotland. The bronze stag-figurine from the temple-pit at Colchester which produced plaques dedicated to Silvanus Callirius has very prominent antlers. The boar-statuettes from Hounslow, the Lexden Tumulus and the temple-site at Muntham Court all have exaggerated dorsal spines.
If we look first at the animal-emphasis, it is apparent that certain aspects have been stressed because of the potent symbolism already latent in that part of the body. The bristle-exaggeration of the boar and the antlers and horns on deer, goats and cattle appear to represent fertility and strength, and as such were obvious features to exaggerate. In the case of the human head, we know that it had special significance for the Celts (infra) and, apart from other considerations, the head is the main means of identifying a man or god and so more care may have been taken over the depiction and emphasis of this crucial part of the body in iconography.
Over-emphasis in Celtic portrayal is interesting in a number of ways. First, negatively, the fertility-bias of Celtic religion does not generally manifest itself in the exaggeration of the sexual/generative parts of the human body. The meaning of the over-emphasis of a human or animal-attribute and the message it was intended to convey to god or devotees have several aspects of which the first is seemingly an acknowledgement or recognition-element. The patron commissioning the sculpture or bronze figure or the craftsman carrying out the manufacture is admitting the power residing in head or horns. To proceed further, the role of exaggeration may be multiple – it may offer special reverence to or propitiation of a particular divinity by means of ‘flattery’ or stress of its power by making the essential part of its body stand out. It may instead (or in addition) remind the worshipper or spectator that the potency of the god is linked to a specific body-part and emanates thence. To my mind an important function of exaggeration is the transmutation from the mundane to the sublime. A convincing parallel may be seen in miniaturisation where model tools and weapons were deliberately produced for sacral purposes. The same, inversely, may be true of exaggeration. To create an image of a beast with extra-large horns was to remove it from the real world to that of the supranatural. A phenomenon related to exaggeration is the addition of extra, often alien, attributes and the consequent representation of a composite being. It is demonstrable that once again the onlooker is being taken from the realism of earth and life to the ‘unreal’ world of the divine, where nature may be contradicted. The ram-horned snake exemplifies such hybridity; the symbolism of both ram and snake are combined to form an image of particular and duplicated potency, rather in the manner of attaching horns to a human being. The association of ram-horned and snake-body on one animal renders the image otherworldly and augments the fertility-symbolism by doubling it as surely as if either snake or ram were repeated – as happens on the Matres sculptures.
The other major composite manifestation, that of horns added to anthropomorphic depictions, is imagery of a very similar nature. We have seen (Chapter Six) that in Celtic cult-iconography beasts were common and acceptable images of the divine (resembling, indeed, their role in Egyptian religion); this was not the case in the classical world. In the Romano-Celtic phase a compromise seems to have been reached in that we have depictions of deities in human form but with horns added. The purpose of this association was presumably to combine the symbolic potency of both types of portrayal. Both man and beast elements were important and both were represented. The sanctity of the iconography appears to have been increased by the presence of composite and unreal imagery, and the association of anthropomorphism and theriomorphism meant that the maximum possible power was vested in the image. The devotee was perhaps visualising his god as an ambiguous mixture of man and beast. In the case of Cernunnos, who is often represented with a stag as well as being antlered himself, one sees the adoption of the animal-attribute perhaps to symbolise the very close and indeed essential rapport between beast and deity.
It is apparent that there are elements common to all the aspects of over-emphasis outlined above. Most important seems to have been the conversion of a ‘normal’ human or animal image to something which, on account of its departure from the realistic, is overtly supernatural or supranatural. For the Celtic suppliant, it was sometimes insufficient to create the image of a god in straightforward human form; it had perhaps to be triplicated, a creature with horns, with a huge head, or a bull with an extra horn. In the case of the three-horned boar, the addition of all three horns is an alien presence, combining the symbolism of the natural ferocity of the boar with the triple horns of a bull whose horned power has itself thus been increased. In all instances of multiplication, addition and exaggeration here examined, the significant act seems to be that of augmentation by visual means. Although by no means always relevant, the lack of an epigraphic tradition in the Celtic world may be a factor. If the devotee was unaccustomed to invoke his god in writing and to inscribe his commitment and devotion, he might have done it visually or iconographically. In this way the worshipper could inform the god that he was aware of his power and, at the same time, he could remind himself and other visitors to shrines of this fact. If it were believed that the god actually resided in the image, then tripling, for instance, of that image would maybe multiply its potency by that factor.
HEADS
The over-emphasis of the head on divine images has been alluded to above. The frequent occurrence of head-representation alone or of ‘severed’ heads is, I believe, part of the same concept in its most extreme form. A number of scholars argue for the presence of a Celtic head-cult, taking as their evidence not only the existence of Celtic head-depictions in stone or metal but also the emphasis given to head-ritual attested both archaeologically by actual skull-collection and by the testimony of Graeco-Roman and Irish literary sources (Chapter One). There is no doubt that the head was considered the most important part of the human body – the emphasis on head-hunting demonstrates this – and the stress on the head in Celtic art is incontestable. Yet I believe it is a mistake to think in terms of a specific head-cult. The significance of the head to the Celts rather means that, as the crucial part, it could on occasions represent the whole. Thus a number of deities could sometimes be depicted by the head alone.
The human head played an important role in continental pre-Roman Celtic art. The Pfalzfeld carved stone pillar of fifth- or fourth-century BC date bears human heads on each of the four sides near the base; a stone head from Mšecké Žehrovice (Czech.) found just outside a viereckshanze was perhaps set up in a shrine as a cult-image by a third- or second-century devotee. The carved piles of heads and the column decorated with incised heads at Entremont may represent something different. The pre-Roman Provençal shrines are unique in presenting abundant and recurring evidence for head-ritual in the form of real skulls (echoing very exactly Graeco-Roman literary sources) and here the sculptures may themselves represent trophies of conquered enemies rather than gods. Likewise, the carved heads beneath the claws of the ‘Tarasque of Noves’ from the same region may represent the triumph of death over human life.
Iron Age continental metalwork abounds in stylised or naturalistic head-symbolism. The Waldalgesheim wine-flagon bears a realistic human face but, by contrast, the head on the Weiskirchen brooch is an integral part of a decorative design. The same duality is present in Britain, though here human representation is much rarer in the Iron Age. Two of the Tal-y-Llyn plaques, found in a first-century BC/AD hoard in Powys, bear two human faces joined by a long neck common to them both, as part of an essentially abstract pattern. But on the Aylesford (Kent) and Marlborough (Wilts.) buckets of first-century BC date, the heads stand out as heads per se, even though stylisation is present.
Most of the numerous depicted heads occurring in Gaul and Britain are of the Romano-Celtic period. The stone examples very frequently have schematised features and it is likely that they portray a number of different gods. The majority of British finds come from the Brigantian hegemony of North Britain where a plethora of local indigenous deities is recorded. Such heads are exemplified by a recent find from Lemington (Tyne & Wear) near Hadrian’s Wall: the head here is a roughly-shaped, irregular rectangle with a mask-like face and coarse features – a thick, moustached mouth, wide flat nose, incised oval eyes and a fringe of hair. Perhaps the most interesting head from outside North Britain is that from Caerwent, Gwent: carved in local stone, it has round open eyes and mask-like features. Its context is particularly illuminating in that it was found on a platform in a chamber which was evidently a shrine, situated in a remote part of the grounds belonging to a late Roman house. Boon has suggested that since the setting is late in the Roman period, the owner of the house may have been Christian while the old Celtic beliefs, perhaps followed by his staff, were banished to a position as far as possible from the house. The presence of a shrine specifically for the head-image argues strongly for its having represented a deity; such an occurrence may be paralleled in Gaul, for instance at the Forêt d’Halatte shrine (Oise) where stone and wooden heads formed a major part of the evidence for cult-activity. Romano-Celtic pots in the form of human heads or faces may also be part of the same tradition. Their sometimes overt function as cremation-receptacles connects them with funerary ritual, but they may on occasions represent deities. In a related context, the presence of tongs on a pot from Colchester may denote a smith-god; a Lincoln face-urn was dedicated unequivocally to Mercury.
The practice of representing a god by the head alone appears to be the result of a complex set of thought-processes. We have seen (Chapter One) that Celtic deities were less bound by functional definition than Graeco-Roman gods; thus overt identification on images was less important. Anonymous anthropomorphic depictions where the whole body is portrayed but without distinctive attributes are recorded at, for example, Redruth and Cirencester. Likewise, deities depicted merely as heads needed no positive physical means of definition: both god and devotee knew who was being invoked. But the head-image had additional properties: set against the proven background of head-ritual, the depiction of the head on its own seems to embody several concepts. It could represent the god pars pro toto: thus the power of the image would have been positively enhanced because the symbolism was concentrated on the part of the body considered most important. Portrayal of the divinity by a head alone may thus have been a deliberate method of honouring a god. Allied to this is the concept of raising the image above the mundane simply because the whole body was not copied from life. The god represented by a head alone could have been as effective a way of de-secularising the image as schematised abstraction, enlargement or multiplication.
MINIATURISATION
The practice of offering miniature replicas of tools, weapons and other ‘secular’ objects to the gods did not originate in the Celtic world nor was it confined to it. The tradition was widespread in antiquity: Near Eastern vehicle-models were made for religious purposes at least as early as 2000 BC and, at the same time, Cretans fashioned as offerings gold and bronze double-axe models. Model agricultural implements were buried in Egyptian tombs (so that the shabti of the deceased could do his work in the afterlife); in Britain, miniature halberd-pendants were being made in the Wessex Early Bronze Age.
Miniature objects appear to have had a particular sanctity in the Celtic world. The presence of pre-Roman models assures their indigenous origin: Late Bronze Age axe- and wheel-models occurring in continental contexts give way in the Iron Age to a much wider variety of models, which are distributed as far apart as Britain and Austria. Axe-models from Long Wittenham, Oxon. and Arras, Yorks. are just two examples from British Iron Age sites. On the other side of the Celtic world, at the oppidum of the Dürrnberg in Austria, several graves contained model objects placed with the dead as apotropaic symbols for the otherworld. The sword and shield and group of three shield-models from Frilford, Oxon. and Worth, Kent are significant in coming specifically from pre-Roman religious sites. It is not until the Romano-Celtic period that we have large-scale evidence for the use of miniature replicas for cult-purposes. In Britain especially, there are numerous axes, spears and other implements which were offered in shrines, as at Woodeaton, Oxon. and Harlow, Essex and many others; in Switzerland axe-models were dedicated to specific deities and have inscriptions on the blades, and in some instances, axe-models bear ritual signs – swastikas, crosses and cosmic motifs.
The precise function of models as votive objects and their symbolism is not at all obvious. The custom of dedicating a small replica of a working implement need only have some such pragmatic purpose as convenience or economy. Certainly, in cases where a model was worn as a talisman, the size-factor is self-evident. But where miniature items were not designed for amuletic use, the question of the role of miniaturisation must be raised. Cost may somtimes be a factor, but there is often more to it than that: model objects are frequently of an expensive metal such as bronze or silver and, more significantly, there is evidence that, in many instances, meticulous care was taken to copy life-size objects in miniature down to the last detail. The recently-discovered axe-model from Tiddington (Warks.) is a case in point: here the craftsman took the trouble to cast the shaft and the blade separately in faithful imitation of a real axe of wood and iron, whilst the simplest technical procedure would have been merely to cast the bronze in one piece. The exact replication of detail is here taken even further in that there is skeuomorphic decoration on the shaft to imitate the wood-grain. If the substitution of a miniature for a full-size object did have a cost element, then it was realised that if it were to be acceptable to the gods, it must be a good copy.
The choice of specific types for miniaturisation could be in part due to the appropriateness of a particular tool to symbolise a particular god or devotee – a soldier might dedicate a sword-model to a war-god, a hunter a spear to a hunter-deity, or a wheel to a sun-god – but the general concept of miniaturisation itself may imply a sacred significance and the faithful replication of objects suggests this. Seen in this light, the practice of model-making can be interpreted as part of the same tradition as exaggeration, plurality and, indeed, ritual breakage which we have seen was a recurrent practice with full-size objects (and indeed model spears at Woodeaton were themselves bent double as a ritual act). Miniaturisation could have been a positive expression of cult and an essential part of ritual. The diminutive size of model objects might actually have enhanced their cult-significance and their potency as symbols. The act of making something too small for normal use could have been a deliberate act of consecration or dedication, in which an item was sanctified through a conscious denial of utility. As with the other types of imagedistortion we have looked at, we may have once again in model objects an example of deliberate removal to the ‘unreal’ world of the supernatural.
Many Celtic images are based upon deliberate ambiguity and double interpretation. La Tène art is described as ‘telling much and concealing much’ (Sandars 1968, 226). In Iron Age metalwork for instance, faces are frequently depicted employing geometric patterns – circles, crescents and other curvilinear forms, with characteristic stress upon pattern. Sometimes it is unclear whether a human face is intended or not and, in any case, there appears to be a deliberate ‘see what you want to see’ element in interpretation. A face can be a face on one level but a set of abstract motifs on another. The literary evidence of classical writers bears this out: ‘in conversation they . . . speak in riddles, for the most part hinting of things and leaving a great deal to be understood’ (Diodorus Siculus V, 31, 1). This deliberate flexibility in interpretation may be seen not only during the Iron Age but later in the Roman period. We have already observed this in the Genii cucullati carving from Cirencester, where the very starkness of the relief may open up its significance to personal interpretation; indeed one could argue that the bare outline only was represented in order to facilitate this flexibility in religious symbolism. A different kind of ambiguity may be seen with some horned beings. For instance, at Uley, Emberton, and on the Smithfield altar in London, Mercury occurs apparently with horns. The London altar is worn and unclear, but at Uley a bronze plaque shows a bust of the god with well-defined horns, and the Emberton stone similarly depicts the messenger-god. But the presence of horns on Mercury-depictions cannot be unequivocal; the classical deity traditionally wears a winged petasos (cap) or has wings sprouting from his hair. In Celto-Roman representation, however, it may be difficult to determine whether horns or wings are intended, and it may be that there is a conscious ambiguity and flexibility here – the worshipper or spectator saw what it was appropriate to him to see – and the god himself, whether horned Celtic deity or Roman winged Mercury, would be content with his image. That such vagueness and substitution may be important is perfectly valid in the context both of the composite symbolism discussed above and in that of the wider issue of schematic and understated expression.
The nature of Celtic religious art, when uncontaminated by the Graeco-Roman artistic and sacral tradition, being often highly stylised, aniconic and abstract, raises the question as to whether the introduction to Celtic lands during the Roman period of classical artistic and religious influences resulted in the demise or, rather, the metamorphosis of Celtic art and religion. Critics of Celtic representation after the Roman occupation accuse indigenous art of having been smothered by the presence of Rome ‘. . . with the gradual advance of Rome a heavy, monumental, rather spiritless classicism engulfs the Continent . . . Rome was moving in and the native arts prepared to go underground’ (Sandars 1968, 233, 273). There are a number of points at issue here: it is recognised that Romano-Celtic stone-carvers were working in an unfamiliar medium, but it has yet to be argued cogently that the results of such superimposition were incompetent. It must also be realised, from a study of Romano-British stone representations – especially in North Britain and the West Country – that a non-Roman sculptural tradition was active and flourishing. Allied to this is the need to acknowledge the continued, albeit altered, existence of Celtic religious cults and concepts. Celtic religion did not die with the Romans. Britain was in the Roman Empire but only just. The gods were, after all Celtic gods. Sulis for example, may have been conflated with Roman Minerva, ‘but her Celtic pedigree remains blazoned for all to see’ (Cunliffe 1979, 158). I would argue, as have others, that the Iron Age abstract Celtic art-tradition continued to some degree after the Roman conquest of Britain. After AD 43 the introduction to Britain of new crafts, such as monumental stone-carving, produced a change of direction in that tradition. In discussing the northern British carvings of a Celtic warrior-deity, Ross comments on the style, admitting iconographic simplicity (in modern or Graeco-Roman terms at least), but acknowledging that individuality and vigour ‘which is immediately impressive reflecting the attempt of local British craftsmen to give visual expression in an unfamiliar medium, to their concept of the local war-god’ (Ross 1967a, 157). There seems to be substantial evidence that whilst Celtic art may have changed direction quite dramatically and radically, it was by no means entirely swamped by the presence of Rome but rather, perhaps, received new stimuli. ‘In Britain, when the artists were of Celtic origin, and had a Celtic stylistic tradition behind them, the use of copy-books did not kill, or even stultify, their native genius’ (Toynbee 1962, 16). Her view, which seems to me to be justified, is that Celtic artists, who possessed a long and distinctive heritage of their own as a backdrop, ‘responded to the challenge of the classical tradition’.
The evidence we have looked at above bears out the argument for the survival of a live Celtic religion and arttradition into the Roman period. Schematic iconography, suggestive of skilful reduction to essentials, the presence of deliberate understatement, ambiguity and interpretive flexibility, and the practice of visual emphasis, all point to a continued flourishing of Celtic religion and its physical expression in art. There existed craftsmen who knew the significance of this iconography and patrons and gods who appreciated these specific qualities. Consideration needs to be given to ability versus choice on the part of Celtic artists. The Welshpool ox-head escutcheon ‘a piece that is as splendidly and bluntly stylised as any representation of a beast could be’ (Toynbee 1962, 17), proclaims survival of late La Tène tradition deep into the period of Roman occupation in Britain. This surely points to the extreme diversity both of artistic capability and British patronal taste. Even where iconographic themes are essentially classical in origin – as in the east Gaulish Jupiter-columns – variations in style, technique and content produce hybrid forms whose originality owes much to the interpretation of indigenous talent; the conflation of ethnic tradition, arising from the interaction of two very different cultures, produced a new and often lively tradition entirely its own, born of the varied response made by British and Gaulish craftsmen to classical forms and themes. The mutual superimposition of Celtic abstract, symbolic and pattern-stressing concepts upon the naturalistic, representational and copy-stressing art of the Graeco-Roman world produced a tradition of physical religious expression that cannot, given its own context and assessed on its own terms, be discussed as inferior art: ‘judged on their own merits and in the light of the aesthetic ideas that shaped them, these . . . carvings can claim to be regarded as the most impressive and original manifestations of art . . .’ (Toynbee 1964, 9).
Hand in hand with Celtic art, this book has sought to demonstrate that Celtic religion was a lively and individual phenomenon during the Romano-Celtic period. More than that, I hope I have shown that this fully-fledged religious expression had deep and complex roots in European prehistory. Whilst it is only in the Roman period that we can name and classify deities, there existed before that time a physically-expressed and complicated series of religious rituals which argue for the Celts’ possession of a deep sense of the natural world and a deliberate desire to come to terms with the vagaries and capricious elements of the supernatural.